Hypothalamic atrophy in progressive supranuclear palsy, assessed by convolutional neural network-based automatic segmentation

Key Points

• To quantify hypothalamic volumes in progressive supranuclear palsy (PSP) using automatic segmentation methods.
• Utilized convolutional neural network (CNN) for brain imaging analysis.
• Compared hypothalamus volumes between PSP patients, controls, and Parkinson's disease (PD) patients.
• Showed significantly reduced hypothalamus volumes in PSP patients compared to controls.
• Findings suggest potential metabolic profiles may correlate with hypothalamic atrophy in PSP.